Related Experiment Video
Updated: Jun 3, 2025
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
Automated segmentation of deep brain structures from Inversion-Recovery MRI
Aigerim Dautkulova1, Omar Ait Aider1, Céline Teulière1
1Université Clermont Auvergne, Clermont Auvergne INP, CNRS, Institut Pascal, F-63000 Clermont-Ferrand, France.
Abstract:
Methods for the automated segmentation of brain structures are a major subject of medical research. The small structures of the deep brain have received scant attention, notably for lack of manual delineations by medical experts. In this study, we assessed an automated segmentation of a novel clinical dataset containing White Matter Attenuated Inversion-Recovery (WAIR) MRI images and five manually segmented structures (substantia nigra (SN), subthalamic nucleus (STN), red nucleus (RN), mammillary body (MB) and mammillothalamic fascicle (MT-fa)) in 53 patients with severe Parkinson's disease. T1 and DTI images were additionally used. We also assessed the reorientation of DTI diffusion vectors with reference to the ACPC line. A state-of-the-art nnU-Net method was trained and tested on subsets of 38 and 15 image datasets respectively. We used Dice similarity coefficient (DSC), 95% Hausdorff distance (95HD), and volumetric similarity (VS) as metrics to evaluate network efficiency in reproducing manual contouring. Random-effects models statistically compared values according to structures, accounting for between- and within-participant variability. Results show that WAIR significantly outperformed T1 for DSC (0.739 ± 0.073), 95HD (1.739 ± 0.398), and VS (0.892 ± 0.044). The DSC values for automated segmentation of MB, RN, SN, STN, and MT-fa decreased in that order, in line with the increasing complexity observed in manual segmentation. Based on training results, the reorientation of DTI vectors improved the automated segmentation.
Insights
Automated brain segmentation using White Matter Attenuated Inversion-Recovery (WAIR) MRI outperformed T1 imaging for deep brain structures in Parkinson
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Computational Neuroscience
Background:
- Automated segmentation of deep brain structures is challenging due to limited manual delineations.
- Accurate segmentation is crucial for understanding neurological disorders like Parkinson's disease.
Purpose of the Study:
- To assess automated segmentation of deep brain structures using White Matter Attenuated Inversion-Recovery (WAIR) MRI.
- To compare WAIR MRI with T1 and DTI imaging for segmenting five key structures in Parkinson's disease patients.
Main Methods:
- A state-of-the-art nnU-Net model was trained and tested on a clinical dataset of 53 Parkinson's patients.
- WAIR, T1, and DTI MRI images were used, along with manual segmentations of substantia nigra (SN), subthalamic nucleus (STN), red nucleus (RN), mammillary body (MB), and mammillothalamic fascicle (MT-fa).
- Evaluation metrics included Dice Similarity Coefficient (DSC), 95% Hausdorff Distance (95HD), and Volumetric Similarity (VS). DTI vector reorientation was also assessed.
Main Results:
- WAIR MRI significantly outperformed T1 MRI across all metrics (DSC, 95HD, VS).
- Segmentation accuracy (DSC) decreased for MB, RN, SN, STN, and MT-fa, correlating with manual segmentation complexity.
- Reorienting DTI diffusion vectors improved automated segmentation performance.
Conclusions:
- WAIR MRI is a superior imaging modality for automated segmentation of deep brain structures compared to T1 MRI.
- The nnU-Net method demonstrates effective segmentation of complex deep brain structures, with performance varying by structure.
- DTI vector reorientation offers a potential enhancement for automated segmentation techniques.
